TF_OLD_DATA ignoring data from the past for frame odom

[plbergeron]

Hi Guys,

We're about to deploy our openptrack system but we're running into new issues at the eleventh hour.

We've got 7 computers: 1 master and 6 imager.

Each imagers are equipped with Zed 1 cameras.

Before the holiday break, we were able to get everything working but we had a lot of trouble calibrating the system because our checkerboard was way too small.
We chose to print a bigger checkerboard that we received after the holidays.
I've updated our camera_network.yaml file to fit with our new checkerboard.
After that, i ran roslaunch opt_calibration opt_calibration_master.launch on the master and roslaunch opt_calibration sensor_<sensor_id>.launch on the nodes (making sure to replace <sensor_id> by the node names).

The first node that we start show no errors.
The error start showing up on the nodes once we start a second node.

i decided to push through anyway, the master sees all the cameras and to calibration windows shows up on the master but the camera can't seem to be able to see the checkerboard.

I came across this post:

https://answers.ros.org/question/201948/tf_old_data-ignoring-data-from-the-past-for-frame-odom/

with this solution the error message doesn't show up but the master won't see the cameras.

Have you guys come across this error before?

[SimeonOA]

Hi,

Can you post an image of the error message you are seeing.

Thank you.

Simeon

[plbergeron]

Sure thing,

[SimeonOA]

Thank you.

From the solution posted, this looks like a time issue.

Did you carry out time synchronization on all the seven computers and verify that the clocks of all OpenPTrack nodes are still synchronized? This should be done before calibration.

[plbergeron]

Yes, i did synchronize the time. The whole network won’t be connected to the internet though so used a different ntp config than the one in the wiki. I was able to get the synchronization working before on this setup though. I’ll try to figure out what’s up with the time synchronization, thanks. Pier-Luc Bergeron Technicien informatique IT Technicien +1 450 272-8122 5333 av. Casgrain, #206, Mtl, QC H2T 1X3 Facebook / Instagram / Linkedin Ce message contient des informations confidentielles. Vous ne pouvez copier, utiliser ou distribuer ces informations si vous n’en êtes pas le destinataire ou autorisé à les recevoir pour le compte du destinataire. ​Si vous avez reçu ce message par erreur, veuillez en informer immédiatement l'expéditeur par téléphone ou le renvoyer immédiatement par courrier électronique. This message contains confidential information. You may not copy, use or distribute this information if you are not the recipient or authorized to receive it on behalf of the recipient. ​If you have received this message in error, please notify immediately the sender at by phone or return it immediately by e-mail. De : Simeon ADEBOLA <notifications@github.com> Envoyé : Tuesday, January 12, 2021 1:14 PM À : OpenPTrack/open_ptrack_v2 <open_ptrack_v2@noreply.github.com> Cc : Pier-Luc Bergeron <plbergeron@normal.studio>; Author <author@noreply.github.com> Objet : Re: [OpenPTrack/open_ptrack_v2] TF_OLD_DATA ignoring data from the past for frame odom (#140) Thank you. From the solution posted, this looks like a time issue. Did you carry out time synchronization<https://github.com/OpenPTrack/open_ptrack_v2/wiki/Time-Synchronization> on all the seven computers and verify<https://github.com/OpenPTrack/open_ptrack_v2/wiki/Time-Synchronization#configuration-2-verification> that the clocks of all OpenPTrack nodes are still synchronized? This should be done before calibration. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub<#140 (comment)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AQ3BM36OETLOV3W3HRHGEPTSZSGNRANCNFSM4V7QOAJA>.

[plbergeron]

So i followed Config 2 on the time synchronization page and i get the same error.